Skip to content Skip to navigation

Shear Thickening in Suspensions

We introduced an experiment to solve the mystry of how Oobleck works.


If you’ve ever baked or cooked, you might be familiar with mixing cornstarch and water. But did you know that when you mix them with equal amount, it is hard to say whether the mixture is a liquid of solid? On the one hand, it does flow (but slowly). On the other hand, this “liquid” can be as stiff as “solid” if you try to stir it vigorously or smash it with a hammer. This phenomena is called shear thickening, in which the fluid viscosity increases with increasing applied force. You can see this effect in the slow motion videos below.



Though people have many cool ideas to play with shear thickening, scientists have been puzzled by the origin of this flow behavior for a long time. Even for a simple thickening mixture of water and micro-particles, there are two competing hypotheses. The most prominent hypothesis suggests that as particles are pushed closer, the resistance from draining the liquid between them locks the particles into clusters (so called “hydrocluster”) that inhibit global flow. More recently however, evidence has been gathering for an alternative mechanism for cluster formation based on particles forming frictional contacts.

In our lab, we investigate shear thickening in colloidal suspensions (simple mixture of water and micro-particles). Instead of cornstarch, we disperse colloidal silica and PMMA (plastic) beads in water-based and oil-based solvents, respectively. We show a SEM image of our silica sphere. These colloidal (small) particles not only exhibit Brownian motion and can be suspended for a long time, their small size (< 10 μm) makes the background liquid “reversible” in motion. This reversible flow, or Stokesian flow, rewinds and follows its previous flow trajectory if you reverse the stir. Imagine that you pushing two colloidal particles closer in a liquid, a part of the resistance you’ll feel is from squeezing the liquid out of the gap between particles. If you pull the particles apart, you’ll feel exactly the same amount of liquid resistance due to the fact that the liquid now needs to flow back into the gap. Furthermore, for Stokesian flow the force required to push the liquid out and suck it back in is the same. This reversible hydrodynamic force is in contrast to the contact force between particles, in which you only feel resistance when you push particles closer. 

We take the advantage of the striking difference between the hydrodynamic and contact forces, and design a flow reversal experiment to settle the debate between these two hypotheses. To do the flow reversal experiment, we need to use a rheometer. In brief, a rheometer consists a lower rotating plate and a upper cone attached to a force sensor. When the lower plate rotates, the “sandwiched”  sample exerts a resistance (shear stress) on the upper cone and can be measured by the force sensor. A schematic image of the rheometer can be found below. In the flow reversal experiment, we first shear the sample until it reaches its steady state, then we suddenly switch the shear direction, maintain the shear speed, and measure the instantaneous response of the sample. By doing so, we can separately determine the hydrodynamic and contact contributions to the response. Because the flow reversibility mentioned previously, the immediate response upon reversal is purely hydrodynamic, and the contact force gradually kicks in a later stage. The prediction is that if hydrodynamic interaction is the dominant role in shear thickening, the magnitude of the suspension stress right before and after the reversal point should be close.

Image from (


In the upper figure , we plot the stress response versus strain (how far the bottom plate moves) after reversal.  The most remarkable feature in this plot is the difference between the sample’s steady response (right plateau) and the response immediately after the switch of shear direction (left plateau). Again, recall that while the right plateau captures the total force (contact + hydro), the left one directly isolates the hydrodynamic term out. The difference between these two plateau is the contact force (total - hydro).

We then repeat this flow reversal procedure at different shear speeds. By doing so, we can observe how contact and hydrodynamic forces evolve as we increase the shear rate and thicken sample. Strikingly, we find that as the suspension viscosity increases, the predominant contribution to this viscosity growth actually comes from the contact force. These results point out that particle contacts play a large role in shear thickening!


In conclusion, we provide the first piece of evidence settling the debate between hydrodynamic and contact pictures. But this experiment is beyond a MythBusters story. Because now we understand that the key to shear thickening is related to the contacts between particles, we can start thinking about how to alter the particle surface to control a shear thickening behavior. This new knob might be helping us to tailor a new shear thickening fluid that perfectly suits the needs in industry and military. For instance, can we condense our commercial product without clogging in the processing pipe? Can we make a fluid that can flow but also stops bullets when it’s shot?